To prewhiten or not to prewhiten in trend analysis?

Abstract The Mann-Kendall test, used to detect a trend in a time series, yields an incorrect (too large) rejection rate when applied to an autocorrelated series with no trend. Prewhitening corrects this situation, but reduces the power of the test when a trend exists. A simulation study is performed to determine when prewhitening can be applied with no real loss of power. It is found that, in general, prewhitening should be avoided when the test has a high power, i.e. when the coefficient of variation is very low, the slope of trend is high, and the sample size is large. In other cases, prewhitening will prevent the false detection of a non-existing trend, without a significant power loss in identifying a trend that exists.

[1]  Khaled H. Hamed,et al.  A modified Mann-Kendall trend test for autocorrelated data , 1998 .

[2]  Richard M. Vogel,et al.  Trends in floods and low flows in the United States: impact of spatial correlation , 2000 .

[3]  Hans von Storch,et al.  Misuses of Statistical Analysis in Climate Research , 1995 .

[4]  S. Yue,et al.  Reply to comment by Xuebin Zhang and Francis W. Zwiers on “Applicability of prewhitening to eliminate the influence of serial correlation on the Mann‐Kendall test” , 2004 .

[5]  Sheng Yue,et al.  The influence of autocorrelation on the ability to detect trend in hydrological series , 2002 .

[6]  S. Yue,et al.  The Mann-Kendall Test Modified by Effective Sample Size to Detect Trend in Serially Correlated Hydrological Series , 2004 .

[7]  S. Yue,et al.  Canadian streamflow trend detection: impacts of serial and cross-correlation , 2003 .

[8]  G. Guo,et al.  Reply to the comment , 1997, quant-ph/9710014.

[9]  Sheng Yue,et al.  Applicability of prewhitening to eliminate the influence of serial correlation on the Mann‐Kendall test , 2002 .

[10]  S. Yue,et al.  Power of the Mann–Kendall and Spearman's rho tests for detecting monotonic trends in hydrological series , 2002 .

[11]  Dennis P. Lettenmaier,et al.  Detection of trends in water quality data from records with dependent observations , 1976 .

[12]  M. Bayazit,et al.  The Power of Statistical Tests for Trend Detection , 2003 .

[13]  Zbigniew W. Kundzewicz,et al.  Change detection in hydrological records—a review of the methodology / Revue méthodologique de la détection de changements dans les chroniques hydrologiques , 2004 .

[14]  Harry F. Lins,et al.  Streamflow trends in the United States , 1999 .

[15]  D. Helsel,et al.  Statistical methods in water resources , 2020, Techniques and Methods.

[16]  Donald H. Burn,et al.  Hydrologic effects of climatic change in west-central Canada , 1994 .

[17]  M. Bayazit,et al.  Comment on “Applicability of prewhitening to eliminate the influence of serial correlation on the Mann‐Kendall test” by Sheng Yue and Chun Yuan Wang , 2004 .

[18]  M. Bayazit,et al.  Comment on: Applicability of prewhitening to eliminate the influence of serial correlation on the Mann-Kendall test. Authors' reply , 2004 .

[19]  S. Yue,et al.  Reply to comment by Mehmetcik Bayazit and Bihrat Önöz on “Applicability of prewhitening to eliminate the influence of serial correlation on the Mann‐Kendall test” , 2004 .

[20]  W. D. Hogg,et al.  Trends in Canadian streamflow , 2000 .

[21]  A. Sankarasubramanian,et al.  Effect of persistence on trend detection via regression , 2003 .